// ⛨ Safety, Evaluation & Security
Automated Red Teaming
Uses one LM to generate test cases that attack another.
自动化红队 2022 Foundational core
IN PLAIN WORDS
A red-team LM generates many prompts likely to elicit harmful or private outputs; a classifier scores success, covering far more than manual testing.
Used in: Pre-deployment safety evaluation
01Original paper
Red Teaming Language Models with Language Models
Ethan Perez, Saffron Huang, Francis Song, et al. · 2022
arXiv ↗ ✓ Verified · arXiv API
02Where it sits
Learn the left column first; the right column builds on this one.
Automated Red Teaming
Then you can learn
- Nothing in the atlas builds on it yet
Related:GCG Universal Jailbreak AttackLLM-as-a-Judge
Left: prerequisites · right: builds on this · dashed: related. Full map →
03Learning resources
Official docs, explainers, courses and reference code. All links checked.
Explainers
04Where the course teaches it
For AI / data
- Verification: verified, checked 2026-10-21.
- JSON: /api/learn/algorithms.json (id:
red-teaming)